Abstract:
In the need of reinforcing the digital improvement, software application transforms into the most
appealing bit of the propelled world. So its common trend to dispense mobile application in such
structure that are isomorphic to the original code. Whereby using the process of reverse
engineering, attackers identify the software behavior and extract the legitimate algorithms where
traditional approaches like firewall, cryptography is not enough. Hence code obfuscation is one
of the indispensable approaches which ratify intensified security by deterring the software code
without changing the semantics and functionality. In this research, we proposed hybrid
methodology to secure the mobile application by using Pro-guard, paranoid and Allatori.
Detailed analysis and comparison of other technologies and their gaps are also part of this
research. We provided an open source package for code obfuscation supporting android. By
using Collberg’s taxonomy we also improved potency, resilience, cost and stealth in mobile
application by using hybrid approach of multiple algorithms. After carrying out test cases, we
evaluated code optimization, apk size shrinking and statement to improve Collberg’s taxonomy.
We successfully obfuscated all java source files. Effective stress test has also been performed to
de-obfuscate the code and unable to get back defined algorithm. Only resource files <xml> are
left to be obfuscate, as main focus of this research was on managed code instead of scripting
code.
